Reorganize a dataframe in by column values R

I have this df

File_pdb Res1 Pos1 Chain1 Res2 Pos2 Chain2
6wps_AH  TRP  105      H  CYS  361      A
6wps_AH  GLU  108      H  LYS  356      A
6wps_AH  PHE  106      H  LYS  356      A
6wps_AH  SER  109      H  THR  345      A
6wps_AH  ILE  111      H  THR  345      A
6wps_AH  TYR  100      H  ASN  343      A

Based on the last two character on the first column 'AH', I-'d like to have as Res1, Pos1 and Chain1 what is referring to 'A' so CYS 361 A. So, the first line would be

File_pdb Res1 Pos1 Chain1 Res2 Pos2 Chain2
6wps_AH  CYS  361      A  TRP  105      H

It should be something general since it can happen that some rows are already right and also, we can have different rows in the same df, some right and others not. Thanks in advance!

here the data:

data.table::as.data.table(structure(list(File_pdb = c("6wps_AH", "6wps_AH", "6wps_AH", 
"6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", 
"6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", "6wps_AH", 
"6wps_AH"), Res1 = c("TRP", "GLU", "PHE", "SER", "ILE", "TYR", 
"GLY", "PHE", "LEU", "TRP", "PHE", "PRO", "ILE", "TRP", "ALA", 
"THR"), Pos1 = c("105", "108", "106", "109", "111", "100", "103", 
"106", "110", "105", "106", "28", "111", "105", "104", "30"), 
    Chain1 = c("H", "H", "H", "H", "H", "H", "H", "H", "H", "H", 
    "H", "H", "H", "H", "H", "H"), Res2 = c("CYS", "LYS", "LYS", 
    "THR", "THR", "ASN", "GLU", "GLU", "GLU", "PRO", "PRO", "LEU", 
    "ASN", "GLU", "GLU", "ASN"), Pos2 = c("361", "356", "356", 
    "345", "345", "343", "340", "340", "340", "337", "337", "335", 
    "343", "340", "340", "334"), Chain2 = c("A", "A", "A", "A", 
    "A", "A", "A", "A", "A", "A", "A", "A", "A", "A", "A", "A"
    )), row.names = c(NA, -16L), class = c("data.table", "data.frame"
)))

Here's a working approach (using data.table) that does what you need. The final column order is not preserved, but that can be handled manually using setcolorder or similar.

I'm using magrittr::%>% for piping here; this is not required and can be removed in favor of data.table-piping ([...][...][...]) for much of it or the use of temporary objects.

library(data.table)
quux %>%
  .[, i := .I] %>%
  melt(id.vars = c("File_pdb", "i"), measure.vars = patterns(Res="^Res", Pos="^Pos", Chain="^Chain")) %>%
  .[, last2 := strsplit(substring(File_pdb, nchar(File_pdb) - 1, nchar(File_pdb)), "")] %>%
  .[, num := mapply(function(nm, l2) match(nm, l2), Chain, last2) ] %>%
  .[, c("variable", "last2") := NULL ] %>%
  dcast(File_pdb + i ~ num, value.var = c("Res", "Pos", "Chain"), sep = "") %>%
  .[, i := NULL] %>%
  .[]
#     File_pdb   Res1   Res2   Pos1   Pos2 Chain1 Chain2
#       <char> <char> <char> <char> <char> <char> <char>
#  1:  6wps_AH    CYS    TRP    361    105      A      H
#  2:  6wps_AH    LYS    GLU    356    108      A      H
#  3:  6wps_AH    LYS    PHE    356    106      A      H
#  4:  6wps_AH    THR    SER    345    109      A      H
#  5:  6wps_AH    THR    ILE    345    111      A      H
#  6:  6wps_AH    ASN    TYR    343    100      A      H
#  7:  6wps_AH    GLU    GLY    340    103      A      H
#  8:  6wps_AH    GLU    PHE    340    106      A      H
#  9:  6wps_AH    GLU    LEU    340    110      A      H
# 10:  6wps_AH    PRO    TRP    337    105      A      H
# 11:  6wps_AH    PRO    PHE    337    106      A      H
# 12:  6wps_AH    LEU    PRO    335     28      A      H
# 13:  6wps_AH    ASN    ILE    343    111      A      H
# 14:  6wps_AH    GLU    TRP    340    105      A      H
# 15:  6wps_AH    GLU    ALA    340    104      A      H
# 16:  6wps_AH    ASN    THR    334     30      A      H

Side note: you might prefer to keep your data in the long (melted) format instead of reshaping back to the wide format. Many analytic processes prefer long (including ggplot2 for plotting), though it does come with some challenges if your toolsets are already set for wide format.